I mean I really don't understand how: you can think that today's llms are going to be the same tomorrow as scaling is exponential, it's non sensical thinking to me and 2) how there is expert consensus on it but still not get it. That's why I made the thread. It doesn't bother me at all, I'm not in the field anymore I made the thread to see how far detached is the tech crowd from what is happening.
> you can think that today's llms are going to be the same tomorrow as scaling is exponential
Irrelevant (even if true, which is far from established); they're still not going to suddenly become capable of reasoning. They will still 'hallucinate' (terrible euphemism, that, but we are where we are).
> how there is expert consensus on it but still not get it.
Only one of the people you mentioned could be considered an expert (Hinton). And he's an expert his a history of very wrong predictions about this sort of thing. There is _by no means_ expert consensus on this.
